A Novel Approach of Integrating Diagenetically Altered Facies with Petrophysical Groups for Static Rock Tying in Lower Cretaceous Carbonate Reservoir of Abu Dhabi, UAE

Abstract

Abstract Carbonate rocks which are born heterogenous are susceptible to diagenetic alterations right after deposition, creating significant heterogeneity in reservoir quality distribution, which makes it challenging to distribute facies and petrophysical properties in the 3D reservoir model with limited data. Therefore, in order to adequately capture the reservoir heterogeneity in the 3D reservoir model a link should be established between the diagenetic facies and petrophysical properties to come up with static rock types (SRTs). The study comprised several integrated steps. The initial step involved a detailed core and thin section description from twenty-nine cored wells to define the lithofacies distribution within the reservoir. Candidate rock types (CRTs) were defined by linking lithofacies to diagenesis. In parallel, petrophysical groups (PGs) were identified from poro-perm and MICP data using machine-learning. SRTs were generated by reconciling the trends and clusters of CRTs and PGs, which were then predicted on un-cored wells using defined boundaries on poro-perm domain. Finally, vertical proportion curves (VPCs) and probability maps for SRTs were created to capture their 3D trends in reservoir model. The depositional setting in the reservoir is dominated by moderate energy mid ramp and moderate to high energy inner ramp facies. Initially eight lithofacies based on rock texture were identified, where the dominant texture is the floatstone rich in Lithocodium/Bacinella (~50%). To initiate the static rock typing workflow, eight lithofacies were split and lumped into seven lithofacies based on dominant grain types and grain vs mud domination. Later these seven lithofacies were split into thirty-one geological facies via integration of lithofacies and diagenesis, which were finally lumped into eight CRTs. On the other hand, eight PGs were identified using machine-learning from poro-perm and MICP data. In the end, eight CRTs were linked with eight PGs through contingency analysis, which showed good relationship to come up with the final eight SRTs. The final SRTs were then designed and optimized to achieve reasonable predictability in un-cored wells, confirmed by the blind tests, which was found to be extremely useful in controlling the distribution of SRTs in the reservoir model. In the 3D reservoir model SRTs were modelled based on the SRTs probability maps and VPCs for each reservoir zone. SRTs were then used to constrain the petrophysical properties and saturation modelling. Properly capturing all the heterogeneities in the reservoir is essential for building a reliable 3D reservoir model that honors the reservoir flow behavior. Strong integration between the geology and petrophysics from the initial steps enabled this SRTs workflow to be executed successfully in order to build a robust 3D reservoir model that captures both geological understanding and flow behavior of the reservoir. Furthermore, this integrated rock typing workflow can be utilized in various carbonate reservoirs.